Stakeholder Consensus on an Interdisciplinary Terminology to Enable the Development and Uptake of Medication Adherence Technologies Across Health Systems: Web-Based Real-Time Delphi Study

Abstract

Background: Technology-mediated medication adherence interventions have proven useful, yet implementation in clinical practice is low. The European Network to Advance Best Practices and Technology on Medication Adherence (ENABLE) European Cooperation in Science and Technology Action (CA19132) online repository of medication adherence technologies (MATechs) aims to provide an open access, searchable knowledge management platform to facilitate innovation and support medication adherence management across health systems. To provide a solid foundation for optimal use and collaboration, the repository requires a shared interdisciplinary terminology.

Objective: We consulted stakeholders on their views and level of agreement with the terminology proposed to inform the ENABLE repository structure.

Methods: A real-time web-based Delphi study was conducted with stakeholders from 39 countries active in research, clinical practice, patient representation, policy making, and technology development. Participants rated terms and definitions of MATech and of 21 attribute clusters on product and provider information, medication adherence descriptors, and evaluation and implementation. Relevance, clarity, and completeness criteria were rated on 9-point scales, and free-text comments were provided interactively. Participants could reconsider their ratings based on real-time aggregated feedback and revisit the survey throughout the study period. We quantified agreement and process indicators for the complete sample and per stakeholder group and performed content analysis on comments. Consensus was considered reached for ratings with a disagreement index of <1. Median ratings guided decisions on whether attributes were considered mandatory, optional, or not relevant. We used the results to improve the terminology and repository structure.

Results: Of 250 stakeholders invited, 117 (46.8%) rated the MATech definition, of whom 83 (70.9%) rated all attributes. Consensus was reached for all items. The definition was considered appropriate and clear (median ratings 7.02, IPR 6.10-7.69, and 7.26, IPR 6.73-7.90, respectively). Most attributes were considered relevant, mandatory, and sufficiently clear to remain unchanged except for ISO certification (considered optional; median relevance rating 6.34, IPR 5.50-7.24) and medication adherence phase, medication adherence measurement, and medication adherence intervention (candidates for optional changes; median clarity ratings 6.07, IPR 4.86-7.17; 6.37, IPR 4.80-6.67; and 5.67, IPR 4.66-6.61, respectively). Subgroup analyses found several attribute clusters considered moderately clear by some stakeholder groups. Results were consistent across stakeholder groups and time, yet response variation was found within some stakeholder groups for selected clusters, suggesting targets for further discussion. Comments highlighted issues for further debate and provided suggestions informing modifications to improve comprehensiveness, relevance, and clarity.

Conclusions: By reaching agreement on a comprehensive MATech terminology developed following state-of-the-art methodology, this study represents a key step in the ENABLE initiative to develop an information architecture capable of structuring and facilitating the development and implementation of MATech across Europe. The debates and challenges highlighted in stakeholders' comments outline a potential road map for further development of the terminology and the ENABLE repository.

International registered report identifier (irrid): RR2-10.1136/bmjopen-2021-059674.

Keywords: Delphi study; behavioral science; digital health; health technology; implementation science; medication adherence; stakeholder engagement.

Figure 1

Median ratings with interpercentile range for relevance and clarity per attribute cluster (N=83). ISO: International Organization for Standardization.

Figure 2

Evolution of median ratings of relevance and clarity per stakeholder group. Each line represents the median value of the ratings for 1 attribute cluster and property (relevance or clarity) calculated on each day of the survey based on the responses of the participants to date. HCP: health care professional; PDM: policy and decision-making.

Figure 3

Screenshot of the European Network to Advance Best Practices and Technology on Medication Adherence (ENABLE) repository as available at manuscript submission date.